Multi-target tracking and interference cooperation method and system based on millimeter wave radar

Abstract

The invention provides a multi-target tracking and interference cooperation method and system based on a millimeter wave radar. The method comprises the following steps: firstly, acquiring an original detection signal in real time by using a millimeter-wave radar carried by an unmanned aerial vehicle, establishing a track correlation sequence of each moving target, extracting interference waveform characteristics formed by signal coupling among a plurality of moving targets in the original detection signal, and then, according to a spatial distribution parameter of the interference waveform characteristics, determining a track correlation sequence of each moving target; correcting the space-time continuity judgment rule of the trajectory correlation sequence, and finally outputting an anti-interference cooperative tracking result based on the corrected space-time continuity judgment rule. According to the technical scheme provided by the invention, the industry pain point that waveform distortion interference of the millimeter wave radar in a dense target scene cannot be eradicated is solved, and the error correlation rate in a dense formation scene is also reduced.

Description

Technical Field

[0001] This application relates to the technical field of UAV radar, and particularly to a multi-target tracking and interference coordination method and system based on millimeter-wave radar. Background Art

[0002] In the field of UAV swarm cooperative operations, such as dense formation flight or obstacle avoidance navigation in complex environments, millimeter-wave radar needs to track multiple high-speed maneuvering targets simultaneously. This scenario requires the system to overcome the mutual coupling interference of radar signals when the target spacing is less than three meters, achieve precise separation of dynamic target echoes, suppress the waveform distortion caused by signal reflection in real time, and ensure the stability of trajectory association across time segments to avoid tracking interruption caused by sudden interference.

[0003] The current mainstream technical solution is based on the fusion method of point cloud clustering and Kalman prediction of FMCW millimeter-wave radar. After generating target point cloud data through a frequency-modulated continuous-wave radar, it uses a density-based spatial clustering algorithm to separate adjacent target point clouds, uses a Kalman filter to predict the target motion state to fill in the trajectory gaps, and presets a fixed threshold to filter out abnormal signal intensity points for interference suppression.

[0004] However, this solution has significant principle defects. Spatial clustering can only separate target positions and cannot eliminate the waveform distortion caused by multiple reflections of electromagnetic waves between targets, resulting in a high mis-association rate of trajectories in dense formations; Kalman prediction overly relies on the historical motion model, resulting in serious trajectory correction delays when sudden maneuvers and transient interferences are superimposed; the static threshold filtering mechanism cannot adapt to dynamic interference changes, causing a large number of interference missed detections. The fundamental crux lies in the split execution of the signal processing and interference suppression links, lacking a real-time closed-loop coordination mechanism. Summary of the Invention

[0005] This application provides a multi-target tracking and interference coordination method and system based on millimeter-wave radar, which is used to solve the problems in the prior art that the waveform distortion interference caused by signal coupling between targets cannot be eliminated, resulting in trajectory mis-association, insufficient adaptability to transient interference causing tracking delay and interruption, and static anti-interference strategies unable to be dynamically adjusted, resulting in too high an interference missed detection rate.

[0006] In a first aspect, this application provides a multi-target tracking and interference coordination method based on millimeter-wave radar, including: Using the millimeter-wave radar carried by the UAV to obtain the original detection signals of multiple moving targets in the flight airspace in real time; Based on the continuous time sequence relationship of the original detection signals, establishing a trajectory association sequence for each moving target; During the generation of the trajectory association sequence, synchronously extracting the interference waveform features formed by signal coupling between multiple moving targets in the original detection signals; According to the spatial distribution parameters of the interference waveform characteristics, the spatio-temporal continuity determination rule of the trajectory association sequence is corrected in real time to obtain the corrected spatio-temporal continuity determination rule; Based on the corrected spatio-temporal continuity determination rule, an anti-interference cooperative tracking result is output.

[0007] Optionally, the original detection signals of multiple moving targets in the flight airspace are obtained in real time by using a millimeter-wave radar carried by a drone, including: During the flight of the drone, the millimeter-wave radar is used to emit millimeter-wave radar scanning signals with continuously changing frequencies at a preset period; Receive the mixed echo signals formed by the reflection of the millimeter-wave radar scanning signals after reaching multiple moving targets in the flight airspace; Perform signal separation operations on the mixed echo signals to obtain reflection signal units independently corresponding to each moving target; According to the flight position parameters of the current drone, map all the reflection signal units to a three-dimensional coordinate system according to the spatial orientation; Output the set of mapped reflection signal units in the three-dimensional coordinate system as the original detection signal.

[0008] Optionally, based on the continuous time sequence relationship of the original detection signals, a trajectory association sequence for each moving target is established, including: For each reflection signal unit, it is decomposed into a set of continuous time segments in the order of acquisition time; Establish three-dimensional space trajectory segments according to the spatial position coordinates and spatial movement trends of the reflection signal units in each time segment set; Between adjacent time segment sets, establish the connection relationship of the three-dimensional space trajectory segments according to the continuity of the spatial position coordinates, the consistency of the spatial movement trends, and the signal strength change law of the reflection signal units; Aggregate the three-dimensional space trajectory segments with connection relationships to form a trajectory association sequence independently corresponding to each moving target.

[0009] Optionally, during the generation process of the trajectory association sequence, the interference waveform characteristics formed by the signal coupling between multiple moving targets in the original detection signals are synchronously extracted, including: When establishing the three-dimensional space trajectory segments, detect the spatial position overlapping areas of multiple reflection signal units within the same time segment; Extract the waveform interaction components of the reflection signal units in the spatial position overlapping areas; Separate the interference waveform segments with morphological deviations from the single-target independent waveforms from the waveform interaction components; Perform temporal aggregation on the interference waveform segments that repeatedly appear in continuous time segments to generate a dynamic interference waveform sequence; Based on the waveform distortion characteristics of the dynamic interference waveform sequence, output interference waveform characteristics including phase continuity and amplitude volatility.

[0010] Optionally, perform temporal aggregation on the interference waveform segments that repeatedly appear in continuous time segments to generate a dynamic interference waveform sequence, including: Establish the continuity identification of interference waveform segments in the overlapping area of the same spatial position on the time axis; According to the time continuity factor between the current time segment and the historical time segment, match the aggregation window of adjacent interference waveform segments; Within the aggregation window, perform a three-dimensional spatial position alignment operation on the interference waveform segments with a time deviation less than a preset threshold; Perform a waveform morphology superposition operation on the interference waveform segments after spatial position alignment to generate an aggregated interference waveform unit; Connect multiple aggregated interference waveform units in the acquisition time order of the time segments to form the dynamic interference waveform sequence.

[0011] Optionally, perform a waveform morphology superposition operation on the interference waveform segments after spatial position alignment to generate an aggregated interference waveform unit, including: Use the interference waveform segments after spatial position alignment as input data to extract a set of waveform data points at the same time order position; Perform a waveform energy superposition operation on all waveform data points in the set of waveform data points to generate a superposition energy distribution; Based on the three-dimensional position distribution of the interference waveform segments after spatial position alignment, calculate the spatial weight factor corresponding to each waveform data point; Use the spatial weight factor to perform an energy equalization operation on the superposition energy distribution to form an equalized waveform data point sequence; Reorganize the equalized waveform data point sequence into a continuous waveform curve in the acquisition time order and output it as an aggregated interference waveform unit according to the continuous waveform curve.

[0012] Optionally, according to the spatial distribution parameters of the interference waveform characteristics, real-time correct the spatio-temporal continuity determination rule of the trajectory association sequence to obtain a corrected spatio-temporal continuity determination rule, including: Map the quantization value of the amplitude volatility in the spatial distribution parameters of the interference waveform characteristics to the spatial position change characteristics of the trajectory association sequence to generate a first rule correction factor; Map the quantization value of the phase continuity in the spatial distribution parameters of the interference waveform characteristics to the motion trend change characteristics of the trajectory association sequence to generate a second rule correction factor; Determine the parameter adjustment strategy of the spatio-temporal continuity determination rule based on the combined relationship between the first rule correction factor and the second rule correction factor; Modify the trajectory position change tolerance threshold and the trajectory direction change tolerance threshold in the spatio-temporal continuity determination rule according to the parameter adjustment strategy; Output the corrected spatio-temporal continuity determination rule including the modified trajectory position change tolerance threshold and the trajectory direction change tolerance threshold.

[0013] In a second aspect, the present application provides a multi-target tracking and interference coordination system based on a millimeter-wave radar, including: An acquisition module, configured to use a millimeter-wave radar carried by a drone to acquire original detection signals of multiple moving targets in a flight airspace in real time; A building module, configured to establish a trajectory association sequence of each moving target based on the continuous time series relationship of the original detection signals; An extraction module, configured to synchronously extract interference waveform features formed by signal coupling between multiple moving targets in the original detection signals during the generation of the trajectory association sequence; A correction module, configured to correct the spatio-temporal continuity determination rule of the trajectory association sequence in real time according to the spatial distribution parameters of the interference waveform features to obtain a corrected spatio-temporal continuity determination rule; An output module, configured to output an anti-interference collaborative tracking result based on the corrected spatio-temporal continuity determination rule.

[0014] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a multi-target tracking and interference coordination method based on a millimeter-wave radar as described in the first aspect above.

[0015] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it implements a multi-target tracking and interference coordination method based on a millimeter-wave radar as described in the first aspect.

[0016] In the embodiments of the present application, by synchronously extracting the interference waveform features during the generation of the trajectory association sequence, a real-time collaborative mechanism for interference feature extraction and target tracking is established, fundamentally eliminating the systematic errors caused by the separation of signal processing and tracking decision-making in traditional methods. The core lies in: accurately capturing the signal coupling area between targets through the detection of the spatial position overlapping area; extracting the waveform interaction components to separate the interference effects of multi-target signals; aggregating the time sequences of interference waveform segments to construct a cross-time interference dynamic evolution model, and finally outputting the quantization features of phase continuity and amplitude volatility, providing a complete physical portrait of dynamic interference for rule modification, and solving the industry pain point that the waveform distortion interference cannot be eradicated in the dense target scenario of millimeter-wave radar.

[0017] Furthermore, a unique double-chain parallel architecture is formed in the technical implementation. By sharing the original detection signal input between the target tracking chain (trajectory association sequence generation) and the interference suppression chain (dynamic interference waveform sequence construction), and real-time interacting at the node of three-dimensional space trajectory segment generation, the interference feature extraction is deeply coupled with the target space motion state, realizing the dynamic collaborative effect between the physical characteristics of interference and the tracking logic rules. And by using the phase continuity feature to quantify the signal stability to directly drive the fault tolerance adjustment of the trajectory direction, and the amplitude volatility feature to map the tolerance correction of the position deviation, the spatio-temporal continuity of trajectory association is maintained in the electromagnetic interference environment, reducing the mis-association rate in the dense formation scenario.

[0018] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0020] Figure 1 Shows a flowchart of a multi-target tracking and interference coordination method based on millimeter-wave radar provided by the present application; Figure 2 Shows a schematic structural diagram of a multi-target tracking and interference coordination system based on millimeter-wave radar provided by the present application; Figure 3 Shows a schematic structural diagram of a computing device provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] In order to enable those skilled in the art to better understand the solutions of the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application.

[0022] In some of the processes described in the specification, claims, and the above-mentioned drawings of this application, a number of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish the different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are of different types.

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present application.

[0024] Figure 1 The following is a flowchart of a multi-target tracking and interference coordination method based on a millimeter-wave radar provided for an embodiment of the present application, as Figure 1 shown, the method includes: Step 101, using the millimeter-wave radar carried by the unmanned aerial vehicle to continuously obtain the original detection signals of multiple moving targets in the flight airspace.

[0025] In this step, the millimeter-wave radar scanning signal refers to a radio beam actively emitted by the unmanned aerial vehicle-mounted radar with a periodically varying frequency, which is used to detect objects in the airspace; the flight airspace is defined as a three-dimensional spatial area within the current flight altitude range of the unmanned aerial vehicle, covering the distribution of obstacles and the dynamic environment of electromagnetic interference; the moving target refers to a collective of other unmanned aerial vehicles or flying objects in the airspace, and its motion state includes position changes and relative motion directions; the mixed echo signal is the superimposed waveform data formed after the scanning signal reaches the moving target and is reflected, and contains the scattering signals of multiple targets; the reflected signal unit refers to the signal subset uniquely corresponding to a single moving target after separation processing, and contains time tags, spatial coordinate vectors, and waveform amplitude characteristics; the original detection signal is defined as the structured data set formed after all the reflected signal units are mapped according to the spatial azimuth, which characterizes the initial perception state of the moving target.

[0026] In this embodiment, first, the millimeter-wave radar controller transmits a scanning signal with linearly varying frequency at a preset period, and the beam covers the real-time flight airspace of the UAV. When the scanning signal reaches a moving target (such as a neighboring UAV), the receiver captures the mixed echo signal formed by its reflection. Then, based on the arrival angle difference and Doppler frequency shift characteristics of the echo signal, a spatial correlation algorithm is used to separate independent reflection signal units (each unit is associated with a single target). Subsequently, by fusing the UAV GNSS positioning data and the radar elevation angle parameter, each reflection signal unit is mapped to the spatial grid nodes of the three-dimensional coordinate system. Finally, all the reflection signal units corresponding to the spatial nodes are integrated to output an original detection signal set containing the target position and motion vector.

[0027] For example, in the scenario of industrial UAV cluster inspection, a host UAV equipped with a millimeter-wave radar flies in the airspace at a height of 50 meters and transmits a continuous frequency modulation wave scanning signal of 77 - 81 GHz. When the signal reaches three auxiliary UAVs (moving targets) 20 meters ahead of it, the receiver obtains the superimposed mixed echo signal. Three independent reflection signal units (corresponding to auxiliary UAVs A / B / C respectively) are separated through beamforming technology, and each unit contains azimuth, distance, and velocity data. Combining the real-time longitude, latitude, and pitch angle of the host UAV, the units are mapped to the northeast celestial coordinate system (three-dimensional coordinate system) to generate an original detection signal set of auxiliary UAVs A(10,5,50), B(12,3,50), and C(15,1,50).

[0028] Step 102: Based on the continuous time sequence relationship of the original detection signals, establish a trajectory association sequence for each moving target.

[0029] In this step, the set of time segments refers to a set of continuous data segments formed by cutting the reflection signal units at a fixed duration; the spatial movement trend represents the direction vector formed by the change in the spatial position of the reflection signal units within a unit time; the three-dimensional space trajectory segment is defined as the vector line segment formed by continuous spatial positions within the same time segment; the connection relationship refers to the association logic that satisfies spatial continuity and motion consistency between adjacent time segment trajectory segments; the trajectory association sequence is a complete description of the target motion path formed by the three-dimensional space trajectory segments aggregated through the connection relationship.

[0030] In this embodiment, first, extract the reflection signal units (such as auxiliary machines A / B / C) in the original detection signal, and divide them into a set of continuous time segments according to the non-uniform cutting strategy based on their timestamps. Secondly, for the reflection signal units of each time segment, extract their three-dimensional position coordinates and calculate the direction vectors formed by adjacent coordinate points to form three-dimensional space trajectory segments with starting / ending coordinates and moving vectors. Then, perform triple determination between the trajectory segments of adjacent time segments: 1) the spatial distance deviation between the end point of the previous trajectory segment and the starting point of the next trajectory segment; 2) the angle deviation between the direction vectors of the two trajectory segments; 3) the difference threshold of the signal intensities of the two trajectory segments. Finally, aggregate the continuous trajectory segments that meet the triple determination to form a structured trajectory association sequence.

[0031] For example, in the UAV inspection scenario following step 101, the time-series positions of the original detection signal containing auxiliary machine A are: t1(10, 5, 50) → t2(10.5, 5.2, 50) → t3(10.8, 5.4, 50). First, cut it into two time segments according to 0.5 seconds: segment 1 (t1 - t2) generates trajectory segment L1 (starting point 10, 5, 50 → ending point 10.5, 5.2, 50, vector 0.5, 0.2, 0); segment 2 (t2 - t3) generates trajectory segment L2 (starting point 10.5, 5.2, 50 → ending point 10.8, 5.4, 50, vector 0.3, 0.2, 0). Secondly, detect that the deviation between the end point of L1 and the starting point of L2 is 0 meters (spatially continuous), the vector angle is 5° (trend consistent), and the signal intensity difference is 3 dB (within the tolerance range), and establish a connection relationship. Finally, output the trajectory association sequence of auxiliary machine A: {L1 → L2}.

[0032] Step 103, during the generation process of the trajectory association sequence, synchronously extract the interference waveform features formed by the signal coupling between multiple moving targets in the original detection signal.

[0033] In this step, the spatial position overlapping area refers to the physical intersection space formed when the three-dimensional coordinate distances of at least two reflection signal units are less than the interference determination threshold within the same time segment; the waveform interaction component represents the synthetic signal feature formed by the mutual interference of the waveform amplitudes and phases of multiple reflection signal units in the overlapping area; the interference waveform segment refers to the distorted signal segment that deviates from the single-target standard waveform due to multi-target coupling in the waveform interaction component; the dynamic interference waveform sequence is an interference signal evolution model formed by aggregating the interference waveform segments of continuous time segments, including the spatial distribution characteristics of waveform distortion.

[0034] In this embodiment, first, while generating the three-dimensional space trajectory segment (output of step 102), the three-dimensional coordinates of the reflection signal units within the same time segment are scanned, and the distances between any two units are calculated. When the distance is less than the interference determination threshold, it is marked as the spatial position overlapping area. Secondly, the original waveform data of multiple reflection signal units within the overlapping area are extracted, and waveform superposition operations are performed to generate waveform interaction components. Then, morphological differential analysis is carried out on the waveform interaction components and a preset single-target independent waveform template, and the segments with waveform distortion exceeding the threshold are extracted as interference waveform segments. Then, the repeated occurrence patterns of the interference waveform segments in consecutive time segments are traced, and a dynamic sequence of the interference waveform segments is constructed in the order of the time axis. Finally, the phase mutation frequency and amplitude oscillation intensity are quantified from the waveform distortion of the dynamic sequence, and the spatial distribution parameters of the interference waveform characteristics are output.

[0035] For example, in the UAV inspection scenario following step 102, the positions of auxiliary machines A and B at time segment t1 are (10, 5, 50) and (9.8, 5.1, 50) respectively, and the distance is 0.25 meters (less than the 1-meter threshold), triggering the spatial position overlapping area; the original waveforms of the two are extracted and superimposed to generate waveform interaction components with enhanced phase oscillation; compared with the single-machine standard waveform template, the amplitude distortion segments (difference > 40%) of this component are identified as interference waveform segments; in three consecutive segments from t1 to t3, this interference waveform segment repeatedly appears in the A - B overlapping area; after aggregation, a dynamic interference waveform sequence is formed, and it is found through analysis that the phase mutation frequency reaches 5 times per second and the amplitude oscillation intensity is ±30%, generating interference waveform characteristic parameters.

[0036] Step 104, according to the spatial distribution parameters of the interference waveform characteristics, the spatio-temporal continuity determination rule of the trajectory association sequence is corrected in real time to obtain the corrected spatio-temporal continuity determination rule.

[0037] In this step, the rule correction factor refers to the continuity determination rule adjustment coefficient generated by converting the interference waveform characteristics, including the position offset tolerance factor and the direction deviation tolerance factor; the parameter adjustment strategy is the dynamic rule control instruction generated according to the combination relationship of the rule correction factors; the trajectory position change tolerance threshold refers to the maximum offset range allowed for the trajectory points in the spatial coordinates; the trajectory direction change tolerance threshold refers to the maximum angular deflection range allowed for the trajectory direction vector within a unit time.

[0038] In this embodiment, first, an amplitude volatility quantization value of the spatial distribution parameter is extracted from the interference waveform features, and it is linearly mapped with the position offset reference value of the trajectory association sequence to generate a position offset fault tolerance factor; at the same time, a phase continuity quantization value is extracted and non-linearly mapped with the trajectory direction deviation reference value to generate a direction deviation fault tolerance factor. Secondly, a parameter adjustment strategy is determined based on the product relationship between the position offset fault tolerance factor and the direction deviation fault tolerance factor: when the product is greater than the dynamic balance critical value, the fault tolerance enhancement mode is activated; when the product is less than the critical value, the standard mode is maintained. Then, in the fault tolerance enhancement mode, the trajectory position change tolerance threshold is expanded to a multiple of the original threshold, and at the same time, the trajectory direction change tolerance threshold is relaxed according to an exponential law; in the standard mode, the original threshold is retained. Finally, the adjusted thresholds are integrated to form a corrected spatio-temporal continuity determination rule including a new spatial position tolerance range and a direction deviation tolerance range.

[0039] For example, in the inspection scenario following step 103, the amplitude volatility quantization value detected in the overlapping area of Auxiliary Machines A - B is 0.7 (reference value 0.5), generating a position offset fault tolerance factor of 1.4; the phase continuity quantization value is 0.6 (reference value 0.8), generating a direction deviation fault tolerance factor of 0.75. The product of the two, 1.05, is greater than the critical value of 1.0, triggering the fault tolerance enhancement mode; first, the trajectory position tolerance threshold is relaxed from 1 meter to 2 meters (position offset fault tolerance factor 1.4 × original threshold 1.4 times), and then the trajectory direction tolerance threshold is relaxed from 15° to 30° (direction deviation fault tolerance factor 0.75 mapped to an exponential function output with 2 times relaxation), and finally, the corrected rule is output: allowing a position offset ≤ 2 meters and a direction deviation ≤ 30°.

[0040] Step 105, output an anti-interference collaborative tracking result based on the corrected spatio-temporal continuity determination rule.

[0041] In this step, the trajectory association verification sequence refers to a structured data set for continuity review of the original trajectory association sequence according to the corrected determination rule; the anti-interference collaborative tracking result is the final output complete tracking path including the spatial coordinates and motion trends of each moving target and meeting the interference suppression conditions.

[0042] In this embodiment, first, extract the original trajectory association sequence generated in step 102 (such as the {L1→L2} sequence of auxiliary machine A), and re-verify the connection state of adjacent three-dimensional space trajectory segments based on the corrected spatio-temporal continuity determination rule (output of step 104): when the spatial deviation between the end point of the trajectory segment and the start point of the next segment is less than the relaxed position tolerance threshold (such as 2 meters), and the movement direction deviation is less than the relaxed direction tolerance threshold (such as 30°), it is determined that the connection is valid. Secondly, perform spatial interpolation smoothing operations on the verified trajectory segments to generate optimized trajectory segments with continuous positions and consistent movement trends. Then, aggregate all the optimized trajectory segments to form a trajectory association verification sequence to eliminate breakpoint anomalies caused by interference. Finally, output the anti-interference collaborative tracking result including the optimized paths of all moving targets.

[0043] For example, in the inspection scenario following step 104 (the distance between the end point (10.5, 5.2, 50) of the trajectory segment L1 of auxiliary machine A and the start point (10.8, 5.4, 50) of L2 is 0.3 meters < 2 meters, and the direction deviation is 10° < 30°), it is determined that the connection is valid; perform linear interpolation on L1-L2 to generate a smooth trajectory; at the same time, the original deviation of L3-L4 of auxiliary machine C due to interference is 2.1 meters (greater than the original 1-meter threshold but less than the new 2-meter threshold) and is re-verified and passed; finally, output the collaborative tracking result: the path of auxiliary machine A is (10, 5, 50) → (10.8, 5.4, 50), and the path of auxiliary machine C is (15, 1, 50) → (17.1, 1.2, 50).

[0044] To sum up, step 101 establishes the three-dimensional space perception basis for multi-target independent reflection signal units; step 102 constructs an anti-interruption three-dimensional trajectory association sequence based on temporal continuity; step 103 synchronously captures signal coupling interference characteristics (including waveform distortion, phase mutation, and amplitude oscillation) during trajectory generation; step 104 dynamically maps the interference characteristics to the determination rule correction factor to achieve real-time adaptive adjustment of the position and direction tolerance thresholds; step 105 re-verifies the trajectory continuity by relaxing the thresholds and outputs the collaborative tracking result that eliminates interference breakpoints. This method realizes the closed-loop collaboration between interference physical characteristics and tracking logic rules for the first time, and maintains the spatio-temporal continuity of trajectory association in signal aliasing scenarios such as UAV formations and urban logistics.

[0045] To solve the problem of distorted target reflection characteristics caused by incomplete separation of mixed echo signals in the prior art, and to improve the spatial positioning accuracy of the original detection signal and the target independence characterization ability, in some embodiments, according to step 101, use the millimeter-wave radar carried by the UAV to obtain the original detection signals of multiple moving targets in the flight airspace in real time, including: Step 201, during the flight of the UAV, use the millimeter-wave radar to emit millimeter-wave radar scanning signals with continuously changing frequencies at a preset period.

[0046] In this step, continuous frequency change refers to the linear increasing or decreasing characteristic of the frequency of the scanning signal within a unit time; the preset period is the signal emission time interval dynamically calculated based on the flight speed of the drone to ensure the refresh requirement of the covered airspace; the millimeter-wave radar scanning signal is a directionally emitted electromagnetic wave beam, and its frequency change range is adaptively determined by the current airspace target detection requirement.

[0047] In this embodiment, first, the drone flight control system calculates the signal emission period according to the real-time flight speed (the faster the speed, the shorter the period), and triggers the millimeter-wave radar controller to generate a scanning signal with continuously changing frequency; the starting frequency of this signal is set based on the maximum distance of the target detected in the previous time, and the frequency change slope is adjusted according to the expected target speed range; subsequently, the signal is emitted by the radar antenna array in a vertical sector scanning mode to cover the three-dimensional space of the forward airspace of the drone.

[0048] Step 202: Receive the mixed echo signal formed after the millimeter-wave radar scanning signal reaches multiple moving targets in the flight airspace and is reflected.

[0049] In this step, the mixed echo signal is defined as a composite signal formed by the superposition of the reflection waveforms of multiple moving targets in the time domain and the frequency domain; reaching a moving target means that the scanning signal forms a reflected wave with the motion characteristics of the target after being scattered by the target surface.

[0050] In this embodiment, first, the millimeter-wave radar receiver array synchronously captures the reflection waveform after the scanning signal reaches the moving target; calculates the arrival angle of the echo signal according to the phase difference of the receiving antenna units; separates the reflection components of targets at different distances by using frequency domain analysis; for the multi-component signals in the same spatial direction, distinguishes the reflected waves of independent targets based on the Doppler frequency shift characteristics, and finally outputs a set of mixed echo signals containing azimuth, distance, and speed parameters.

[0051] Step 203: Perform signal separation operation on the mixed echo signal to obtain a reflection signal unit independently corresponding to each moving target.

[0052] In this step, the signal separation operation refers to the processing process of decoupling the mixed echo signal into independent target signals; the reflection signal unit is defined as the signal data packet uniquely corresponding to a single moving target after separation, containing the distance, azimuth, speed parameters of the target and the original waveform characteristics.

[0053] In this embodiment, first, perform joint time-frequency domain analysis on the mixed echo signal to extract component signals in different spatial directions; second, establish a signal independence discrimination criterion based on the Doppler frequency shift differences of each component signal; then, perform phase unwrapping on the signal components that meet the independence criterion to eliminate multi-target signal interference; finally, assign a unique target identifier to each decoupled signal and package it into a reflection signal unit containing distance, azimuth, velocity, and waveform.

[0054] Step 204: According to the flight position parameters of the current unmanned aerial vehicle (UAV), map all reflection signal units to a three-dimensional coordinate system according to the spatial azimuth.

[0055] In this step, the flight position parameters include the real-time positioning coordinates (longitude, latitude, altitude) of the UAV and the three-dimensional attitude angles (pitch angle, roll angle, yaw angle); the spatial azimuth mapping refers to the calculation process of converting the polar coordinate data (azimuth angle, distance) of the reflection signal unit into the absolute coordinates of the three-dimensional geodetic coordinate system; the three-dimensional coordinate system adopts the Earth-Centered Earth-Fixed (ECEF) coordinate system, and its grid reference plane coincides with the standard ellipsoid.

[0056] In this embodiment, first, obtain the six-degree-of-freedom pose parameters in real time through the UAV navigation system: the longitude λ, latitude φ, and altitude h output by the GPS, and the pitch angle θ, roll angle ψ, and yaw angle ω output by the Inertial Measurement Unit (IMU); second, perform geometric transformation on the azimuth angle α and distance d of each reflection signal unit: in the first step, calculate the local rectangular coordinates (forward x / transverse y / vertical z) of the target in the UAV body coordinate system; in the second step, apply the coordinate rotation matrix to process the carrier attitude angle; in the third step, perform coordinate transformation to obtain the ECEF coordinates (X, Y, Z) of the target; finally, establish a mapping data table containing the three-dimensional coordinates of all targets.

[0057] Step 205: Output the set of reflection signal units mapped in the three-dimensional coordinate system as the original detection signal.

[0058] In this step, the set of reflection signal units after mapping is a structured data set after three-dimensional coordinate transformation; the original detection signal is defined as a complete perception data packet containing the spatial positions, motion characteristics, and waveform attributes of each moving target.

[0059] In this embodiment, first, all target coordinates in the three-dimensional mapping data table are integrated, and a unified data structure is constructed with the motion parameters (velocity vector, signal strength) and waveform features extracted in step 203. Secondly, a timestamp index is established according to the target ID to ensure temporal continuity. Then, the rationality of the spatial position is verified: abnormal units with a height exceeding the airspace range (>2000 meters) or a relative speed exceeding the limit (>100 m / s) are eliminated. Finally, it is encapsulated and output in a standard data format, including fields: target ID, timestamp, ECEF coordinates (X, Y, Z), velocity vector (Vx, Vy, Vz), signal strength value, and waveform segment.

[0060] In order to solve the problems of cross-time trajectory association breakage and loss of sudden maneuvering targets in the prior art, and to improve the spatio-temporal continuity guarantee ability of three-dimensional space trajectories, in some embodiments, according to step 102, based on the continuous temporal relationship of the original detection signals, a trajectory association sequence of each moving target is established, including: Step 301, for each reflection signal unit, it is decomposed into a set of continuous time segments in the order of acquisition time.

[0061] In this step, the set of time segments refers to the discrete data segments obtained by dividing the continuous time series data of the reflection signal unit according to a fixed duration or event trigger condition; the acquisition time order maintains the chronological relationship of the original detection timestamps of the reflection signal units.

[0062] In this embodiment, first, according to the timestamp information carried by the reflection signal unit, a time window is generated with a preset duration division strategy. Secondly, the reflection signal units with continuous timestamps within the window are aggregated into independent time segments. Then, time alignment correction is performed on the boundary data of each time segment to eliminate the temporal breakage caused by clock jitter. Finally, a linked list of time segments arranged in chronological order is constructed to ensure that adjacent segments are temporally continuous and non-overlapping.

[0063] Step 302, a three-dimensional space trajectory segment is established according to the spatial position coordinates and spatial movement trend of the reflection signal units in each set of time segments.

[0064] In this step, the spatial movement trend represents the direction vector of the change in the spatial position of the reflection signal unit within the time segment; the three-dimensional space trajectory segment is defined as the vector line segment formed by the starting and ending positions of the time segment, with an additional momentum direction attribute.

[0065] Step 302: A three-dimensional space trajectory segment is established according to the spatial position coordinates and spatial movement trend of the reflection signal units in each set of time segments.

[0066] In this embodiment, first, the spatial position coordinates of the first and last frame reflection signal units within the time segment are extracted and defined as the starting coordinate and the ending coordinate respectively. Secondly, a direction vector of the spatial movement trend is established based on the geometric relationship from the starting coordinate to the ending coordinate, and this vector describes the movement direction and spatial displacement relationship of the target within the time segment. Then, the spatial distance distribution analysis is performed on the spatial position coordinates of all reflection signal units within this time segment and the connection line between the starting and ending points to generate a position consistency evaluation value. Finally, when the position consistency evaluation value meets the set requirements, a three-dimensional space trajectory segment including the starting coordinate, the ending coordinate, and the direction vector is created.

[0067] Step 303: Between adjacent time segment sets, establish the connection relationship of the three-dimensional space trajectory segments according to the continuity of the spatial position coordinates of the reflection signal units, the consistency of the spatial movement trend, and the signal intensity change law.

[0068] In this step, the continuity of the spatial position coordinates refers to the range of distance deviation between the ending point of the previous trajectory segment and the starting point in the three-dimensional coordinate system; the consistency of the spatial movement trend characterizes the angular similarity between the direction vector of the previous segment and the direction vector of the next segment; the signal intensity change law describes the gradual change characteristic of the reflection intensity of the same target in the cross-time connection.

[0069] In this embodiment, first, detect the spatial deviation amount between the ending coordinate of the last three-dimensional space trajectory segment in the previous time segment set and the starting coordinate of the first trajectory segment in the next time segment set. Secondly, compare the angular differences of the three-dimensional components of the direction vectors of the two trajectory segments and calculate the angular offset amount. Then, analyze the change gradient of the signal intensity before and after the connection point to detect whether it conforms to the natural attenuation or enhancement law of the target reflection intensity. Then, when the spatial deviation amount is within the set distance range, the angular offset amount is within the set angular range, and the signal intensity change is monotonically gradual, establish a strong connection relationship; if some conditions do not match, establish a weak connection relationship. Finally, output a connection relationship diagram with connection strength markings.

[0070] Step 304: Aggregate the three-dimensional space trajectory segments with connection relationships to form a trajectory association sequence corresponding independently to each moving target.

[0071] In this step, the aggregation operation refers to the process of linking the trajectory segments that meet the connection conditions in chronological order; the trajectory association sequence is a complete description of the target movement path verified by continuity and consistency.

[0072] In this embodiment, first, select pairs of trajectory segments with connection strength exceeding a set threshold in the connection relationship diagram; second, connect the head and tail of the strongly connected trajectory segments in chronological order to form an initial path chain; then, insert spatial interpolation trajectory segments for the weakly connected trajectory segments for transitional connection; then, assign a unique target identifier to each initial path chain; finally, verify the spatio-temporal continuity of the path chain: if there is a missing trajectory segment, complete it based on the motion trend prediction, and output a trajectory association sequence including the target ID, the trajectory segment sequence, and the motion trend.

[0073] To solve the problem of incomplete and static extraction of waveform features of signal coupling interference in the prior art, and to improve the dynamic collaborative perception accuracy of interference features and target motion states, in some embodiments, as described in step 103, during the generation process of the trajectory association sequence, interference waveform features formed by signal coupling between multiple moving targets in the original detection signal are synchronously extracted, including: Step 401, when establishing three-dimensional space trajectory segments, detect the spatial position overlapping area of multiple reflection signal units within the same time segment.

[0074] In this step, the spatial position overlapping area refers to the physical intersection space formed when the three-dimensional coordinate distance between two or more reflection signal units within the same time segment is less than a set spatial threshold; the set spatial threshold is jointly determined by the beam width of the airborne millimeter-wave radar and the target size characteristics.

[0075] In this embodiment, first, within the same time period when generating three-dimensional space trajectory segments (such as step 302), traverse all reflection signal units in the current time segment; second, calculate the Euclidean space distance between any two units in the three-dimensional coordinate system; then, when the distance value is less than the dynamic threshold calculated based on the radar wavelength and the minimum target spacing, mark this area as the spatial position overlapping area; finally, record the identification of all reflection signal units included in the overlapping area and their common intersection coordinates.

[0076] Step 402, extract the waveform interaction components of the reflection signal units in the spatial position overlapping area.

[0077] In this step, the waveform interaction component refers to the synthetic waveform feature formed by the mutual interference of the original waveforms of multiple reflection signal units within the spatial position overlapping area; the original waveform is defined by the time-domain amplitude and phase data carried by the reflection signal unit.

[0078] In this embodiment, first, the original waveform data of each reflection signal unit within the spatially overlapping region is acquired; second, after aligning the waveforms of all units within the region along the time axis, amplitude superposition and phase coupling operations are performed; then, the local extreme value features of the superimposed waveform are extracted: including the amplitude peak position, phase mutation points, and time-frequency distribution characteristics; finally, a waveform interaction component containing the following elements is generated: the time-domain graph of the superimposed waveform, the phase coherence graph, and the time-frequency energy distribution characteristics.

[0079] Step 403: Separate from the waveform interaction component the interference waveform segments that have a morphological deviation from the independent waveform of a single target.

[0080] In this step, the interference waveform segment is defined as the local waveform interval within the waveform interaction component where its morphology deviates from the preset single-target waveform template due to the coupling of multi-target signals; the morphological deviation includes abnormal amplitude oscillations, phase jumps, and unnatural increases or decreases in frequency components.

[0081] In this embodiment, first, a single-target reference waveform reference set is established based on the radar scattering physical model, and this reference set contains the standard reflection waveform characteristics of different components of the unmanned aerial vehicle under interference-free conditions. Then, the time-domain amplitude variation, phase continuity, and frequency-domain energy distribution of the waveform interaction component are compared with the reference waveform in multiple dimensions: detecting the mutation interval where the amplitude value exceeds the standard floating range, identifying the discontinuous jump points that do not match the standard phase change curve, and locating the frequency bands with abnormal energy aggregation in the frequency domain. Then, the interference waveform segments are determined according to the preset persistence condition, that is, when there are at least two types of significant deviations (such as amplitude mutation accompanied by phase jump) in a certain interval and the duration exceeds the minimum threshold, the start and end points of the time of this section and the frequency boundary are intercepted, and finally, a structured interference waveform segment containing elements such as the time window and deviation type is generated.

[0082] Step 404: Aggregate the interference waveform segments that repeatedly appear in consecutive time segments in time sequence to generate a dynamic interference waveform sequence.

[0083] In this step, time-sequence aggregation means concatenating the interference waveform segments in the same spatially overlapping region along the time axis; the dynamic interference waveform sequence is a cross-time model containing the interference evolution characteristics, and its core parameters include the waveform distortion propagation direction and the intensity change gradient.

[0084] Step 405: Based on the waveform distortion characteristics of the dynamic interference waveform sequence, output the interference waveform characteristics including phase continuity and amplitude volatility.

[0085] In this embodiment, first, interference source association is established through spatial position continuity: the central coordinates of the overlapping regions corresponding to the interference waveform segments are extracted, and when the coordinate offset between consecutive segments is less than the radar spatial resolution, it is determined as homologous interference. Then, cross-time feature evolution analysis is performed: the time sequence is calibrated based on the starting time of the first segment, the co-directional change trend of the amplitude deviation between adjacent segments (such as monotonically increasing amplitude) is analyzed, the propagation characteristics of the phase jump direction (such as clockwise rotation trend) are traced, and the migration vector of the central coordinates (such as continuous offset towards the northeast direction) is calculated. Then, a dynamic interference model is comprehensively constructed: the amplitude change gradient, phase propagation law, and spatial migration direction are integrated to form an interference evolution path with spatio-temporal continuity. Finally, a dynamic interference waveform sequence that completely describes the interference evolution law is output.

[0086] To solve the problem in the prior art that discrete interference segments cannot characterize the cross-time interference evolution law, and to improve the spatio-temporal correlation modeling ability of dynamic interference waveforms, in some embodiments, according to what is described in step 404, the interference waveform segments that repeatedly appear in consecutive time segments are temporally aggregated to generate a dynamic interference waveform sequence, including: Step 501, establish a continuity identifier on the time axis for the interference waveform segments in the overlapping region of the same spatial position.

[0087] In this step, the continuity identifier refers to an order association mark established between interference waveform segments in the time dimension, used to characterize the continuous attribute of interference events with the same spatial origin evolving over time; the time axis refers to the sequence structure of interference waveform segments arranged in chronological order.

[0088] In this embodiment, first, the interference waveform segments are grouped based on the coordinate hash value of the spatial position overlapping region. After sorting the segments within each group in chronological order, it is detected whether the time interval between adjacent segments meets the signal propagation continuity requirement: that is, the difference between the end time of the previous segment and the start time of the next segment needs to be less than the maximum propagation delay time of the radar signal. At the same time, the feature inheritance between segments is analyzed, including whether the amplitude change trends are co-directional and whether the phase jump patterns are similar. When both the time continuity and feature similarity conditions are met, a unique continuity identifier is added to the segment sequence of this group.

[0089] Step 502, match the aggregation window of adjacent interference waveform segments according to the time continuity factor between the current time segment and the historical time segment.

[0090] In this step, the time continuity factor is a comprehensive index that quantifies the strength of the temporal correlation between segments; the aggregation window refers to a set of analysis time periods for interference events divided based on the time continuity factor.

[0091] In this embodiment, first, the continuity identifier of the current interference waveform segment and its historical sequence data are extracted, and the time continuity factor is calculated: the time interval value between the current segment and the most recent historical segment is obtained and converted into an interval score, and at the same time, the segment feature similarity score is extracted, and the time continuity factor value is generated through weighted fusion. If the factor value is higher than the dynamic aggregation threshold, the current segment is included in the aggregation window of the historical segments; if the factor does not meet the standard but the spatial coordinate deviation is within the positioning accuracy range, a new sub-window is created and a spatial association chain is established with the main window. Finally, an aggregation window description body including the time span, the core spatial coordinates, and the evolution characteristics is generated.

[0092] Step 503, within the aggregation window, perform a three-dimensional spatial position alignment operation on the interference waveform segments with a time deviation less than a preset threshold.

[0093] In this step, the three-dimensional spatial position alignment operation refers to the process of compensating the spatial positions of the interference waveform segments with slight time offsets, eliminating the position parsing error caused by signal propagation delay.

[0094] In this embodiment, first, the spatial position information and timestamps of all interference waveform segments within the aggregation window are obtained. A position compensation model is constructed based on the consistent characteristic of the propagation speed of electromagnetic waves in the air: the position coordinates of each segment are corrected for time offset (i.e., the spatial offset is scaled according to the time deviation ratio), so that the actual spatial positions of all segments are mapped to the same reference time. Then, position normalization processing is performed: the corrected coordinates are uniformly converted to the origin position of the core spatial coordinate system of the aggregation window, and finally, a set of time-aligned spatial positions is generated.

[0095] Step 504, perform a waveform morphology superposition operation on the interference waveform segments after spatial position alignment to generate an aggregated interference waveform unit.

[0096] In this step, the waveform morphology superposition operation refers to the process of fusing multiple waveform characteristics in the same spatial coordinate system.

[0097] In this embodiment, first, the amplitude-phase data sequences of each interference waveform segment after spatial alignment are extracted. A data point mapping relationship is established through precise matching on the time axis: the amplitude and phase values are sampled at the same relative time nodes. Subsequently, weighted superposition calculation is performed: weights are assigned according to the distance between the spatial position of each segment and the origin of the core coordinate system (the closer the distance, the higher the weight). The distorted points with prominent characteristics are retained during the superposition process: such as the peak positions of amplitude mutations and the discontinuous points of phase jumps. Finally, an aggregated interference waveform unit including the integrated amplitude curve and typical distortion feature markers is formed.

[0098] Step 505, connect multiple aggregated interference waveform units in the acquisition time order of the time segments to form the dynamic interference waveform sequence.

[0099] In this embodiment, first, all aggregated interference waveform units are sorted in the chronological order of the aggregation window. An evolutionary logic connection between units is established: the migration continuity of adjacent units in characteristic parameters is detected (such as the gradient of the gradual change of the amplitude peak value and the consistency of the phase jump direction). Transition markers are inserted at positions with logical breakpoints, and at the same time, the characteristic evolution trend is recorded (such as the continuous increase of the amplitude peak value or the diffusion of the phase jump angle). Finally, a complete sequence including the time dimension, spatial distribution, and characteristic evolution map is constructed.

[0100] To solve the problem of aggregation distortion caused by ignoring the spatial attenuation characteristics in waveform superposition in the prior art, and to improve the energy distribution balance and physical authenticity of interference waveform units, in some embodiments, according to what is described in step 504, the interference waveform segments that repeatedly appear in continuous time segments are aggregated in time sequence to generate a dynamic interference waveform sequence, including: Step 601: Using the interference waveform segments aligned by the spatial position as input data, a set of waveform data points at the same time sequence position is extracted.

[0101] In this step, the set of waveform data points is defined as a synchronous sampling data cluster of all interference waveform segments at the same relative time node, including multi-dimensional parameters such as amplitude values, phase angles, and frequency components.

[0102] In this embodiment, first, a unified time reference axis is established: taking the starting moment of the aggregation window as zero, timestamp normalization processing is performed on all interference waveform segments aligned in space. Subsequently, sampling points are set at fixed intervals on the normalized time axis, and the amplitude values, phase angles, and spectral energy distribution values of each segment at the same time node are extracted. For non-integer sampling points, data interpolation between segments is used to generate a continuous waveform curve. Finally, a set of waveform data points including time node encoding and corresponding multi-segment data values is constructed.

[0103] Step 602: Perform a waveform energy superposition operation on all waveform data points in the set of waveform data points to generate a superposition energy distribution.

[0104] Step 601: Using the interference waveform segments aligned by the spatial position as input data, a set of waveform data points at the same time sequence position is extracted.

[0105] In this step, the set of waveform data points is defined as a synchronous sampling data cluster of all interference waveform segments at the same relative time node, including multi-dimensional parameters such as amplitude values, phase angles, and frequency components.

[0106] In this embodiment, first, a unified time reference axis is established: taking the starting moment of the aggregation window as zero, timestamp normalization processing is performed on all spatially aligned interference waveform segments. Subsequently, sampling points are set at fixed intervals on the normalized time axis, and the amplitude values, phase angles, and spectral energy distribution values of each segment at the same time node are extracted. For non-integer sampling points, data interpolation between segments is used to generate a continuous waveform curve. Finally, a set of waveform data points including time node encoding and corresponding multi-segment data values is constructed.

[0107] Step 602: Perform a waveform energy superposition operation on all waveform data points in the set of waveform data points to generate a superposition energy distribution.

[0108] In this step, the waveform energy superposition operation refers to the process of fusing the energy characteristics of multi-source data points to generate a synthetic energy field; the superposition energy distribution characterizes the integrated form of the comprehensive energy in the time-frequency-space dimension.

[0109] In this embodiment, first, each waveform data point is parsed into a three-dimensional energy vector, and then energy fusion is performed: for all data points at the same time node, the average value of the time-domain vector modulus length, the weighted sum value of the frequency-domain vector (the weight is determined by the signal propagation attenuation model), and the geometric center position of the spatial domain vector are calculated respectively. Finally, it is integrated into a superposition energy distribution body including a synthetic time-domain energy curve, a fused frequency-domain spectral feature, and a core spatial energy point.

[0110] Step 603, based on the three-dimensional position distribution of the interference waveform segments aligned by the spatial position, calculate the spatial weight factor corresponding to each waveform data point.

[0111] In this step, the spatial weight factor is a proportional coefficient that quantifies the importance of the spatial position of the waveform data point and is used to correct the spatial attenuation effect during the energy superposition of multi-source signals; the three-dimensional position distribution refers to the spatial coordinate cluster of all interference waveform segments after the alignment operation.

[0112] In this embodiment, the specific implementation process first establishes a spatial position influence model: taking the core coordinates of the aggregation window as the center origin, calculate the radial distance value of the three-dimensional position corresponding to each waveform data point. According to the energy attenuation law of electromagnetic waves propagating in free space, a weight distribution function is constructed: the closer the position is to the core origin, the higher the weight, and the farther the distance, the weight decays exponentially. At the same time, density analysis is performed on the cluster positions: weight equalization adjustment is performed in the spatially dense area to prevent overcompensation, and finally, a unique spatial weight factor value corresponding to each waveform data point is generated, which comprehensively combines the distance factor and the position density factor.

[0113] Step 604, use the spatial weight factor to perform an energy equalization operation on the superposition energy distribution to form an equalized waveform data point sequence.

[0114] In this step, the energy equalization operation refers to the process of correcting the superimposed energy distribution using the spatial weight factor; the equalized waveform data point sequence is a standardized waveform feature data set after eliminating the spatial position deviation.

[0115] In this embodiment, first, the time-domain energy value, frequency-domain energy value, and spatial energy point coordinates in the superimposed energy distribution are extracted. Three-dimensional correction is performed using the spatial weight factor. Among them, the time-domain correction is the time-domain energy value divided by the corresponding spatial weight factor, the frequency-domain correction is the frequency-domain energy value multiplied by the position density factor, and the spatial reconstruction is that the spatial energy point coordinates contract towards the core origin according to the weight ratio; then waveform feature reconstruction is performed, and the standard waveform amplitude curve and phase change curve are regenerated based on the corrected energy values. Finally, a data sequence in which the time nodes and the equalized waveform features are strictly matched is constructed.

[0116] Step 605, reorganize the equalized waveform data point sequence into a continuous waveform curve in the order of acquisition time, and output it as an aggregated interference waveform unit according to the continuous waveform curve.

[0117] In this step, the continuous waveform curve is a smooth waveform path formed by connecting the equalized waveform data points along the time axis, representing the time-domain evolution form of the interference signal after energy equalization; the aggregated interference waveform unit is defined as a standardized interference output module that includes complete waveform features and spatial correlation attributes.

[0118] In this embodiment, first, the time node encoding of the equalized waveform data point sequence is read in time order, and the corrected amplitude value, phase angle, and spatial correction coordinates of each node are extracted. The discrete data points are concatenated using the spline path connection algorithm: a coordinate system is established with the amplitude value as the vertical axis variable and the time node as the horizontal axis, and a non-linear interpolation operation is performed on the data points to generate a continuous amplitude change curve; at the same time, the phase angle is mapped to an independent coordinate system to generate a continuous phase change curve. Subsequently, feature points of the two curves are marked: an energy density mark is added at the amplitude peak, and a discontinuity mark is added at the phase jump point. Finally, the amplitude curve, phase curve, and spatial coordinate evolution path are integrated to form an aggregated interference waveform unit.

[0119] To solve the problem that the determination rule in the prior art is static and cannot adapt to dynamic interference changes, and to improve the real-time collaborative optimization ability of the spatio-temporal continuity determination threshold and interference characteristics, in some embodiments, according to what is described in step 104, the spatio-temporal continuity determination rule of the trajectory association sequence is corrected in real time according to the spatial distribution parameters of the interference waveform characteristics, and the corrected spatio-temporal continuity determination rule is obtained, including: Step 701, map the quantization value of the amplitude volatility in the spatial distribution parameters of the interference waveform characteristics to the spatial position change characteristics of the trajectory association sequence to generate a first rule correction factor.

[0120] In this step, the spatial position change feature characterizes the position offset fluctuation degree of consecutive trajectory points in the trajectory association sequence; the first rule correction factor refers to the position tolerance threshold adjustment coefficient generated based on the amplitude fluctuation characteristic.

[0121] In this embodiment, first, the quantization value of the amplitude volatility in the interference waveform feature (this value is calculated from the dynamic interference waveform sequence) is extracted, and at the same time, the spatial position change feature of the trajectory association sequence (including the eigenvalue distribution of the variance matrix of the three-dimensional coordinates) is obtained. By establishing a non-linear mapping model: the amplitude fluctuation value is input into the radial basis function network, and the network output is subjected to matrix dot multiplication with the eigenvalue of the position change feature to generate an initial correction weight matrix. Finally, principal component analysis is performed on this matrix, and the component value in the direction of the maximum eigenvector is normalized to the interval from 0 to 1, and the first rule correction factor is output. The value of this factor is positively correlated with the amplitude fluctuation intensity.

[0122] Step 702, map the quantization value of the phase continuity in the spatial distribution parameters of the interference waveform feature to the motion trend change feature of the trajectory association sequence to generate a second rule correction factor.

[0123] In this step, the motion trend change feature describes the angular fluctuation characteristic of the direction vector of the trajectory association sequence; the second rule correction factor is the trajectory direction tolerance adjustment coefficient generated based on the phase continuity.

[0124] In this embodiment, first, the quantization value of the phase continuity (reflecting the phase mutation frequency of the interference waveform) is fused with the motion trend change feature (including the direction vector angle variance and the high-order derivative). A chaotic mapping model is constructed: the phase continuity value is used as the initial value parameter to iteratively generate a chaotic sequence, and this sequence is subjected to convolution operation with the high-order derivative of the motion trend feature. The amplitude of the fundamental frequency component of the convolution result is extracted through Fourier spectrum analysis and compressed to the range from 0 to 1 by the logistic function. Finally, the second rule correction factor is output, and the factor value is negatively correlated with the phase continuity.

[0125] Step 703, determine the parameter adjustment strategy of the spatio-temporal continuity determination rule based on the combination relationship between the first rule correction factor and the second rule correction factor.

[0126] In this step, the combination relationship refers to the weight distribution and linkage mechanism between the two correction factors; the parameter adjustment strategy is a threshold control instruction set generated according to the dynamic relationship between the two factors.

[0127] In this embodiment, a two-factor collaborative analysis model is first established: perform dynamic weight distribution operations on the first rule correction factor and the second rule correction factor to generate a dominant factor identifier (position priority / direction priority / balanced mode). When the first factor is significantly higher than the second factor, activate the position tolerance threshold dominant mode; when the second factor is significantly dominant, activate the direction tolerance threshold dominant mode; when the difference between the two factors is within the collaborative range, trigger the balanced adjustment mode. Subsequently, set the threshold adjustment coefficient matrix according to the mode type: the position dominant mode uses an exponential amplification coefficient, the direction dominant mode uses a polynomial update coefficient, and the balanced mode uses a linear expansion coefficient. Finally, output a parameter adjustment strategy instruction set including the mode identifier, adjustment function type, and reference coefficient.

[0128] Step 704, modify the trajectory position change tolerance threshold and the trajectory direction change tolerance threshold in the spatio-temporal continuity determination rule according to the parameter adjustment strategy.

[0129] In this step, the trajectory position change tolerance threshold allows the maximum offset range of the trajectory point in the three-dimensional space; the trajectory direction change tolerance threshold allows the maximum angular deflection of the trajectory direction vector.

[0130] In this embodiment, first read the threshold base value in the original determination rule: the position tolerance threshold base value is determined by the radar positioning accuracy, and the direction tolerance threshold base value is preset according to the target maneuverability. Select the corresponding adjustment function according to the parameter adjustment strategy: input the reference coefficient into the adjustment function (linear / exponential / polynomial) to generate an adjustment coefficient matrix. Then perform threshold dimension update: multiply the position tolerance threshold base value by the position dimension adjustment coefficient, and multiply the direction tolerance threshold base value by the direction dimension adjustment coefficient to obtain the new threshold value. Finally, add dynamic safety margin compensation: automatically scale the threshold range according to the size of the reference coefficient (widen the margin when the coefficient is large / narrow the margin when the coefficient is small), and output the corrected determination rule including the final threshold.

[0131] Step 705, output the corrected spatio-temporal continuity determination rule including the modified trajectory position change tolerance threshold and the trajectory direction change tolerance threshold.

[0132] In this step, the corrected spatio-temporal continuity determination rule is a trajectory association verification criterion that has been dynamically calibrated. Its core parameters are the updated spatial position offset tolerance range and the motion direction deflection tolerance range; the rule output structure includes the threshold application condition, the effective time window, and the exception handling protocol.

[0133] In this embodiment, first, integrate the new threshold values generated in step 704 with the original rule framework: update the trajectory position change tolerance threshold to a spherical tolerance domain that includes the upper limit of the three-dimensional space coordinate deviation (the diameter is defined by the new position threshold), and at the same time reconstruct the trajectory direction change tolerance threshold into a conical space that includes the combined constraints of the pitch angle and the yaw angle. Then inject dynamic constraint conditions: set the time interval and spatial range for the threshold to take effect (such as only taking effect in the sector where signal interference is detected), and bind the exception handling rule (trigger spatial interpolation instead of disconnection when the threshold is exceeded). Finally, encapsulate it into a machine-executable instruction set, which includes a four-dimensional structure of threshold parameters, spatial activation domain, time scope, and exception handling logic.

[0134] Figure 2 FIG. provides a schematic structural diagram of a multi-target tracking and interference coordination system based on a millimeter-wave radar according to an embodiment of the present application, as Figure 2 shown, the system includes: An acquisition module 21, configured to use a millimeter-wave radar carried by a drone to acquire original detection signals of multiple moving targets in the flight airspace in real time; A building module 22, configured to establish a trajectory association sequence of each moving target based on the continuous time sequence relationship of the original detection signals; [[ID=?]] An extraction module 23, configured to synchronously extract, during the generation of the trajectory association sequence, interference waveform features formed by signal coupling between multiple moving targets in the original detection signals; A correction module 24, configured to correct the spatio-temporal continuity determination rule of the trajectory association sequence in real time according to the spatial distribution parameters of the interference waveform features, to obtain a corrected spatio-temporal continuity determination rule; An output module 25, configured to output an anti-interference collaborative tracking result based on the corrected spatio-temporal continuity determination rule.

[0135] Figure 2 The multi-target tracking and interference coordination system based on a millimeter-wave radar can execute Figure 1 The multi-target tracking and interference coordination method based on a millimeter-wave radar described in the embodiment shown, the implementation principle and technical effects will not be elaborated. For the multi-target tracking and interference coordination system based on a millimeter-wave radar in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0136] In a possible design, Figure 2 The multi-target tracking and interference coordination system based on a millimeter-wave radar described in the embodiment shown can be implemented as a computing device, as Figure 3 shown, the computing device may include a storage component 31 and a processing component 32; It should be noted that there seems to be a missing number in the "ID=?" line. Please check and correct if necessary.The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32.

[0137] The processing component 32 is used for the above Figure 1 A multi-target tracking and interference coordination method based on a millimeter-wave radar according to the embodiment.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A multi-target tracking and interference coordination method based on millimeter-wave radar, characterized in that Including: Using a millimeter-wave radar carried by a drone to obtain the original detection signals of multiple moving targets in the flight airspace in real time; Based on the continuous time-series relationship of the original detection signals, establishing a trajectory association sequence for each moving target; During the generation process of the trajectory association sequence, synchronously extracting the interference waveform features formed by signal coupling between multiple moving targets within the original detection signals; According to the spatial distribution parameters of the interference waveform features, real-time correcting the spatio-temporal continuity determination rule of the trajectory association sequence to obtain the corrected spatio-temporal continuity determination rule; Outputting an anti-interference collaborative tracking result based on the corrected spatio-temporal continuity determination rule.

2. The method according to claim 1, wherein Using a millimeter-wave radar carried by a drone to obtain the original detection signals of multiple moving targets in the flight airspace in real time, including: During the flight of the drone, using the millimeter-wave radar to emit millimeter-wave radar scanning signals with continuously changing frequencies at a preset period; Receiving the mixed echo signals formed by the reflection of the millimeter-wave radar scanning signals after reaching multiple moving targets in the flight airspace; Performing signal separation operations on the mixed echo signals to obtain reflection signal units independently corresponding to each moving target; According to the flight position parameters of the current drone, mapping all the reflection signal units to a three-dimensional coordinate system according to the spatial azimuth; Outputting the set of mapped reflection signal units in the three-dimensional coordinate system as the original detection signals.

3. The method according to claim 1, wherein Based on the continuous time-series relationship of the original detection signals, establishing a trajectory association sequence for each moving target, including: For each reflection signal unit, decomposing it into a set of continuous time segments in the order of acquisition time; Establishing three-dimensional space trajectory segments according to the spatial position coordinates and spatial movement trends of the reflection signal units in each time segment set; Between adjacent time segment sets, establishing the connection relationship of the three-dimensional space trajectory segments according to the continuity of the spatial position coordinates, the consistency of the spatial movement trends, and the signal intensity change law of the reflection signal units; Aggregating the three-dimensional space trajectory segments with connection relationships to form a trajectory association sequence independently corresponding to each moving target.

4. The method according to claim 1, wherein During the generation process of the trajectory association sequence, synchronously extracting the interference waveform features formed by signal coupling between multiple moving targets within the original detection signals, including: When establishing three-dimensional space trajectory segments, detecting the spatial position overlapping regions of multiple reflection signal units within the same time segment; Extracting the waveform interaction components of the reflection signal units in the spatial position overlapping regions; Separating the interference waveform segments with morphological deviations from the single-target independent waveforms from the waveform interaction components; Performing temporal aggregation on the interference waveform segments repeatedly appearing in continuous time segments to generate a dynamic interference waveform sequence; Based on the waveform distortion features of the dynamic interference waveform sequence, outputting the interference waveform features including phase continuity and amplitude volatility.

5. The method according to claim 4, characterized in that, Performing temporal aggregation on the interference waveform segments repeatedly appearing in continuous time segments to generate a dynamic interference waveform sequence, including: Establishing the continuity identification of the interference waveform segments in the same spatial position overlapping region on the time axis; According to the time continuity factor between the current time segment and the historical time segment, matching the aggregation window of adjacent interference waveform segments; Within the aggregation window, perform a three-dimensional spatial position alignment operation on interference waveform segments with a time deviation less than a preset threshold; Perform a waveform morphology superposition operation on the interference waveform segments after spatial position alignment to generate an aggregated interference waveform unit; Connect multiple aggregated interference waveform units in the acquisition time order of the time segments to form the dynamic interference waveform sequence.

6. The method according to claim 5, characterized in that Performing a waveform morphology superposition operation on the interference waveform segments after spatial position alignment to generate an aggregated interference waveform unit includes: Using the interference waveform segments after spatial position alignment as input data, extract a set of waveform data points at the same time order position; Perform a waveform energy superposition operation on all waveform data points in the set of waveform data points to generate a superposition energy distribution; Based on the three-dimensional position distribution of the interference waveform segments after spatial position alignment, calculate the spatial weight factor corresponding to each waveform data point; Use the spatial weight factor to perform an energy equalization operation on the superposition energy distribution to form an equalized waveform data point sequence; Reorganize the equalized waveform data point sequence into a continuous waveform curve in the acquisition time order and output it as an aggregated interference waveform unit according to the continuous waveform curve.

7. The method according to claim 1, wherein According to the spatial distribution parameters of the interference waveform characteristics, real-time correct the spatio-temporal continuity determination rule of the trajectory association sequence to obtain a corrected spatio-temporal continuity determination rule, including: Map the quantization value of the amplitude volatility in the spatial distribution parameters of the interference waveform characteristics to the spatial position change characteristics of the trajectory association sequence to generate a first rule correction factor; Map the quantization value of the phase continuity in the spatial distribution parameters of the interference waveform characteristics to the motion trend change characteristics of the trajectory association sequence to generate a second rule correction factor; Based on the combined relationship between the first rule correction factor and the second rule correction factor, determine the parameter adjustment strategy of the spatio-temporal continuity determination rule; Modify the trajectory position change tolerance threshold and the trajectory direction change tolerance threshold in the spatio-temporal continuity determination rule according to the parameter adjustment strategy; Output a corrected spatio-temporal continuity determination rule including the modified trajectory position change tolerance threshold and the trajectory direction change tolerance threshold.

8. A multi-target tracking and interference coordination system based on millimeter-wave radar, characterized in that, Including: An acquisition module for using a millimeter-wave radar carried by a drone to acquire the original detection signals of multiple moving targets in the flight airspace in real time; A building module for establishing a trajectory association sequence of each moving target based on the continuous time series relationship of the original detection signals; An extraction module for synchronously extracting the interference waveform characteristics formed by signal coupling between multiple moving targets in the original detection signals during the generation of the trajectory association sequence; A correction module for real-time correcting the spatio-temporal continuity determination rule of the trajectory association sequence according to the spatial distribution parameters of the interference waveform characteristics to obtain a corrected spatio-temporal continuity determination rule; An output module for outputting an anti-interference collaborative tracking result based on the corrected spatio-temporal continuity determination rule.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a multi-target tracking and interference coordination method based on a millimeter-wave radar as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, A computer program is stored, and when the computer program is executed by a computer, it implements a multi-target tracking and interference coordination method based on a millimeter-wave radar as described in any one of claims 1 to 7.

Images